diff options
author | Jesús <heckyel@hyperbola.info> | 2022-05-17 10:10:39 +0800 |
---|---|---|
committer | Jesús <heckyel@hyperbola.info> | 2022-05-17 10:10:39 +0800 |
commit | 4bbf329feb5a820ac21269fa426c95ca14d7af25 (patch) | |
tree | 2c147a162b4bddc7862ed5895f1f66edd9a675e8 /yt_dlp/YoutubeDL.py | |
parent | e21342911839b7796a5c788a7c3f13b06d975c64 (diff) | |
parent | 5faf6528fb701724ac32e0a487f92281c7800bda (diff) | |
download | hypervideo-pre-4bbf329feb5a820ac21269fa426c95ca14d7af25.tar.lz hypervideo-pre-4bbf329feb5a820ac21269fa426c95ca14d7af25.tar.xz hypervideo-pre-4bbf329feb5a820ac21269fa426c95ca14d7af25.zip |
updated from upstream | 17/05/2022 at 10:10
Diffstat (limited to 'yt_dlp/YoutubeDL.py')
-rw-r--r-- | yt_dlp/YoutubeDL.py | 402 |
1 files changed, 196 insertions, 206 deletions
diff --git a/yt_dlp/YoutubeDL.py b/yt_dlp/YoutubeDL.py index ab7cf3d6d..94f8dcaef 100644 --- a/yt_dlp/YoutubeDL.py +++ b/yt_dlp/YoutubeDL.py @@ -1,8 +1,4 @@ #!/usr/bin/env python3 -# coding: utf-8 - -from __future__ import absolute_import, unicode_literals - import collections import contextlib import datetime @@ -16,6 +12,7 @@ import locale import operator import os import platform +import random import re import shutil import subprocess @@ -24,97 +21,115 @@ import tempfile import time import tokenize import traceback -import random import unicodedata - -from enum import Enum +import urllib.request from string import ascii_letters +from .cache import Cache from .compat import ( - compat_basestring, - compat_brotli, compat_get_terminal_size, - compat_kwargs, - compat_numeric_types, compat_os_name, - compat_pycrypto_AES, compat_shlex_quote, compat_str, - compat_tokenize_tokenize, compat_urllib_error, compat_urllib_request, - compat_urllib_request_DataHandler, windows_enable_vt_mode, ) from .cookies import load_cookies +from .downloader import FFmpegFD, get_suitable_downloader, shorten_protocol_name +from .downloader.rtmp import rtmpdump_version +from .extractor import _LAZY_LOADER +from .extractor import _PLUGIN_CLASSES as plugin_extractors +from .extractor import gen_extractor_classes, get_info_extractor +from .extractor.openload import PhantomJSwrapper +from .minicurses import format_text +from .postprocessor import _PLUGIN_CLASSES as plugin_postprocessors +from .postprocessor import ( + EmbedThumbnailPP, + FFmpegFixupDuplicateMoovPP, + FFmpegFixupDurationPP, + FFmpegFixupM3u8PP, + FFmpegFixupM4aPP, + FFmpegFixupStretchedPP, + FFmpegFixupTimestampPP, + FFmpegMergerPP, + FFmpegPostProcessor, + MoveFilesAfterDownloadPP, + get_postprocessor, +) from .utils import ( + DEFAULT_OUTTMPL, + LINK_TEMPLATES, + NO_DEFAULT, + NUMBER_RE, + OUTTMPL_TYPES, + POSTPROCESS_WHEN, + STR_FORMAT_RE_TMPL, + STR_FORMAT_TYPES, + ContentTooShortError, + DateRange, + DownloadCancelled, + DownloadError, + EntryNotInPlaylist, + ExistingVideoReached, + ExtractorError, + GeoRestrictedError, + HEADRequest, + InAdvancePagedList, + ISO3166Utils, + LazyList, + MaxDownloadsReached, + Namespace, + PagedList, + PerRequestProxyHandler, + Popen, + PostProcessingError, + ReExtractInfo, + RejectedVideoReached, + SameFileError, + UnavailableVideoError, + YoutubeDLCookieProcessor, + YoutubeDLHandler, + YoutubeDLRedirectHandler, age_restricted, args_to_str, - ContentTooShortError, date_from_str, - DateRange, - DEFAULT_OUTTMPL, determine_ext, determine_protocol, - DownloadCancelled, - DownloadError, encode_compat_str, encodeFilename, - EntryNotInPlaylist, error_to_compat_str, - ExistingVideoReached, expand_path, - ExtractorError, filter_dict, float_or_none, format_bytes, - format_field, format_decimal_suffix, + format_field, formatSeconds, - GeoRestrictedError, get_domain, - has_certifi, - HEADRequest, - InAdvancePagedList, int_or_none, iri_to_uri, - ISO3166Utils, join_nonempty, - LazyList, - LINK_TEMPLATES, locked_file, make_dir, make_HTTPS_handler, - MaxDownloadsReached, merge_headers, network_exceptions, - NO_DEFAULT, number_of_digits, orderedSet, - OUTTMPL_TYPES, - PagedList, parse_filesize, - PerRequestProxyHandler, platform_name, - Popen, - POSTPROCESS_WHEN, - PostProcessingError, preferredencoding, prepend_extension, - ReExtractInfo, register_socks_protocols, - RejectedVideoReached, remove_terminal_sequences, render_table, replace_extension, - SameFileError, sanitize_filename, sanitize_path, sanitize_url, sanitized_Request, std_headers, - STR_FORMAT_RE_TMPL, - STR_FORMAT_TYPES, str_or_none, strftime_or_none, subtitles_filename, @@ -123,52 +138,19 @@ from .utils import ( to_high_limit_path, traverse_obj, try_get, - UnavailableVideoError, url_basename, variadic, version_tuple, write_json_file, write_string, - YoutubeDLCookieProcessor, - YoutubeDLHandler, - YoutubeDLRedirectHandler, ) -from .cache import Cache -from .minicurses import format_text -from .extractor import ( - gen_extractor_classes, - get_info_extractor, - _LAZY_LOADER, - _PLUGIN_CLASSES as plugin_extractors -) -from .extractor.openload import PhantomJSwrapper -from .downloader import ( - FFmpegFD, - get_suitable_downloader, - shorten_protocol_name -) -from .downloader.rtmp import rtmpdump_version -from .postprocessor import ( - get_postprocessor, - EmbedThumbnailPP, - FFmpegFixupDuplicateMoovPP, - FFmpegFixupDurationPP, - FFmpegFixupM3u8PP, - FFmpegFixupM4aPP, - FFmpegFixupStretchedPP, - FFmpegFixupTimestampPP, - FFmpegMergerPP, - FFmpegPostProcessor, - MoveFilesAfterDownloadPP, - _PLUGIN_CLASSES as plugin_postprocessors -) -from .version import __version__ +from .version import RELEASE_GIT_HEAD, __version__ if compat_os_name == 'nt': import ctypes -class YoutubeDL(object): +class YoutubeDL: """YoutubeDL class. YoutubeDL objects are the ones responsible of downloading the @@ -329,13 +311,17 @@ class YoutubeDL(object): has been filtered out. break_per_url: Whether break_on_reject and break_on_existing should act on each input URL as opposed to for the entire queue - cookiefile: File name where cookies should be read from and dumped to + cookiefile: File name or text stream from where cookies should be read and dumped to cookiesfrombrowser: A tuple containing the name of the browser, the profile name/pathfrom where cookies are loaded, and the name of the keyring. Eg: ('chrome', ) or ('vivaldi', 'default', 'BASICTEXT') legacyserverconnect: Explicitly allow HTTPS connection to servers that do not support RFC 5746 secure renegotiation nocheckcertificate: Do not verify SSL certificates + client_certificate: Path to client certificate file in PEM format. May include the private key + client_certificate_key: Path to private key file for client certificate + client_certificate_password: Password for client certificate private key, if encrypted. + If not provided and the key is encrypted, yt-dlp will ask interactively prefer_insecure: Use HTTP instead of HTTPS to retrieve information. At the moment, this is only supported by YouTube. http_headers: A dictionary of custom headers to be used for all requests @@ -427,10 +413,14 @@ class YoutubeDL(object): sleep_interval_subtitles: Number of seconds to sleep before each subtitle download listformats: Print an overview of available video formats and exit. list_thumbnails: Print a table of all thumbnails and exit. - match_filter: A function that gets called with the info_dict of - every video. - If it returns a message, the video is ignored. - If it returns None, the video is downloaded. + match_filter: A function that gets called for every video with the signature + (info_dict, *, incomplete: bool) -> Optional[str] + For backward compatibility with youtube-dl, the signature + (info_dict) -> Optional[str] is also allowed. + - If it returns a message, the video is ignored. + - If it returns None, the video is downloaded. + - If it returns utils.NO_DEFAULT, the user is interactively + asked whether to download the video. match_filter_func in utils.py is one example for this. no_color: Do not emit color codes in output. geo_bypass: Bypass geographic restriction via faking X-Forwarded-For @@ -504,7 +494,7 @@ class YoutubeDL(object): care about HLS. (only for youtube) """ - _NUMERIC_FIELDS = set(( + _NUMERIC_FIELDS = { 'width', 'height', 'tbr', 'abr', 'asr', 'vbr', 'fps', 'filesize', 'filesize_approx', 'timestamp', 'release_timestamp', 'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count', @@ -512,7 +502,7 @@ class YoutubeDL(object): 'start_time', 'end_time', 'chapter_number', 'season_number', 'episode_number', 'track_number', 'disc_number', 'release_year', - )) + } _format_fields = { # NB: Keep in sync with the docstring of extractor/common.py @@ -579,7 +569,7 @@ class YoutubeDL(object): def check_deprecated(param, option, suggestion): if self.params.get(param) is not None: - self.report_warning('%s is deprecated. Use %s instead' % (option, suggestion)) + self.report_warning(f'{option} is deprecated. Use {suggestion} instead') return True return False @@ -681,7 +671,7 @@ class YoutubeDL(object): pp_def = dict(pp_def_raw) when = pp_def.pop('when', 'post_process') self.add_post_processor( - get_postprocessor(pp_def.pop('key'))(self, **compat_kwargs(pp_def)), + get_postprocessor(pp_def.pop('key'))(self, **pp_def), when=when) self._setup_opener() @@ -696,7 +686,7 @@ class YoutubeDL(object): with locked_file(fn, 'r', encoding='utf-8') as archive_file: for line in archive_file: self.archive.add(line.strip()) - except IOError as ioe: + except OSError as ioe: if ioe.errno != errno.ENOENT: raise return False @@ -782,9 +772,9 @@ class YoutubeDL(object): assert hasattr(self, '_output_process') assert isinstance(message, compat_str) line_count = message.count('\n') + 1 - self._output_process.stdin.write((message + '\n').encode('utf-8')) + self._output_process.stdin.write((message + '\n').encode()) self._output_process.stdin.flush() - res = ''.join(self._output_channel.readline().decode('utf-8') + res = ''.join(self._output_channel.readline().decode() for _ in range(line_count)) return res[:-len('\n')] @@ -893,16 +883,19 @@ class YoutubeDL(object): raise DownloadError(message, exc_info) self._download_retcode = 1 - class Styles(Enum): - HEADERS = 'yellow' - EMPHASIS = 'light blue' - ID = 'green' - DELIM = 'blue' - ERROR = 'red' - WARNING = 'yellow' - SUPPRESS = 'light black' + Styles = Namespace( + HEADERS='yellow', + EMPHASIS='light blue', + FILENAME='green', + ID='green', + DELIM='blue', + ERROR='red', + WARNING='yellow', + SUPPRESS='light black', + ) def _format_text(self, handle, allow_colors, text, f, fallback=None, *, test_encoding=False): + text = str(text) if test_encoding: original_text = text # handle.encoding can be None. See https://github.com/yt-dlp/yt-dlp/issues/2711 @@ -910,8 +903,6 @@ class YoutubeDL(object): text = text.encode(encoding, 'ignore').decode(encoding) if fallback is not None and text != original_text: text = fallback - if isinstance(f, self.Styles): - f = f.value return format_text(text, f) if allow_colors else text if fallback is None else fallback def _format_screen(self, *args, **kwargs): @@ -972,7 +963,7 @@ class YoutubeDL(object): self.to_screen('Deleting existing file') def raise_no_formats(self, info, forced=False, *, msg=None): - has_drm = info.get('__has_drm') + has_drm = info.get('_has_drm') ignored, expected = self.params.get('ignore_no_formats_error'), bool(msg) msg = msg or has_drm and 'This video is DRM protected' or 'No video formats found!' if forced or not ignored: @@ -993,11 +984,9 @@ class YoutubeDL(object): outtmpl_dict.update({ k: sanitize(v) for k, v in DEFAULT_OUTTMPL.items() if outtmpl_dict.get(k) is None}) - for key, val in outtmpl_dict.items(): + for _, val in outtmpl_dict.items(): if isinstance(val, bytes): - self.report_warning( - 'Parameter outtmpl is bytes, but should be a unicode string. ' - 'Put from __future__ import unicode_literals at the top of your code file or consider switching to Python 3.x.') + self.report_warning('Parameter outtmpl is bytes, but should be a unicode string') return outtmpl_dict def get_output_path(self, dir_type='', filename=None): @@ -1007,12 +996,6 @@ class YoutubeDL(object): expand_path(paths.get('home', '').strip()), expand_path(paths.get(dir_type, '').strip()) if dir_type else '', filename or '') - - # Temporary fix for #4787 - # 'Treat' all problem characters by passing filename through preferredencoding - # to workaround encoding issues with subprocess on python2 @ Windows - if sys.version_info < (3, 0) and sys.platform == 'win32': - path = encodeFilename(path, True).decode(preferredencoding()) return sanitize_path(path, force=self.params.get('windowsfilenames')) @staticmethod @@ -1022,7 +1005,7 @@ class YoutubeDL(object): # '%%' intact for template dict substitution step. Working around # with boundary-alike separator hack. sep = ''.join([random.choice(ascii_letters) for _ in range(32)]) - outtmpl = outtmpl.replace('%%', '%{0}%'.format(sep)).replace('$$', '${0}$'.format(sep)) + outtmpl = outtmpl.replace('%%', f'%{sep}%').replace('$$', f'${sep}$') # outtmpl should be expand_path'ed before template dict substitution # because meta fields may contain env variables we don't want to @@ -1070,7 +1053,7 @@ class YoutubeDL(object): formatSeconds(info_dict['duration'], '-' if sanitize else ':') if info_dict.get('duration', None) is not None else None) - info_dict['autonumber'] = self.params.get('autonumber_start', 1) - 1 + self._num_downloads + info_dict['autonumber'] = int(self.params.get('autonumber_start', 1) - 1 + self._num_downloads) info_dict['video_autonumber'] = self._num_videos if info_dict.get('resolution') is None: info_dict['resolution'] = self.format_resolution(info_dict, default=None) @@ -1078,7 +1061,7 @@ class YoutubeDL(object): # For fields playlist_index, playlist_autonumber and autonumber convert all occurrences # of %(field)s to %(field)0Nd for backward compatibility field_size_compat_map = { - 'playlist_index': number_of_digits(info_dict.get('_last_playlist_index') or 0), + 'playlist_index': number_of_digits(info_dict.get('__last_playlist_index') or 0), 'playlist_autonumber': number_of_digits(info_dict.get('n_entries') or 0), 'autonumber': self.params.get('autonumber_size') or 5, } @@ -1092,18 +1075,18 @@ class YoutubeDL(object): # Field is of the form key1.key2... # where keys (except first) can be string, int or slice FIELD_RE = r'\w*(?:\.(?:\w+|{num}|{num}?(?::{num}?){{1,2}}))*'.format(num=r'(?:-?\d+)') - MATH_FIELD_RE = r'''(?:{field}|{num})'''.format(field=FIELD_RE, num=r'-?\d+(?:.\d+)?') + MATH_FIELD_RE = rf'(?:{FIELD_RE}|-?{NUMBER_RE})' MATH_OPERATORS_RE = r'(?:%s)' % '|'.join(map(re.escape, MATH_FUNCTIONS.keys())) - INTERNAL_FORMAT_RE = re.compile(r'''(?x) + INTERNAL_FORMAT_RE = re.compile(rf'''(?x) (?P<negate>-)? - (?P<fields>{field}) - (?P<maths>(?:{math_op}{math_field})*) + (?P<fields>{FIELD_RE}) + (?P<maths>(?:{MATH_OPERATORS_RE}{MATH_FIELD_RE})*) (?:>(?P<strf_format>.+?))? (?P<remaining> (?P<alternate>(?<!\\),[^|&)]+)? (?:&(?P<replacement>.*?))? (?:\|(?P<default>.*?))? - )$'''.format(field=FIELD_RE, math_op=MATH_OPERATORS_RE, math_field=MATH_FIELD_RE)) + )$''') def _traverse_infodict(k): k = k.split('.') @@ -1182,7 +1165,7 @@ class YoutubeDL(object): fmt = outer_mobj.group('format') if fmt == 's' and value is not None and key in field_size_compat_map.keys(): - fmt = '0{:d}d'.format(field_size_compat_map[key]) + fmt = f'0{field_size_compat_map[key]:d}d' value = default if value is None else value if replacement is None else replacement @@ -1197,7 +1180,7 @@ class YoutubeDL(object): value = map(str, variadic(value) if '#' in flags else [value]) value, fmt = ' '.join(map(compat_shlex_quote, value)), str_fmt elif fmt[-1] == 'B': # bytes - value = f'%{str_fmt}'.encode('utf-8') % str(value).encode('utf-8') + value = f'%{str_fmt}'.encode() % str(value).encode() value, fmt = value.decode('utf-8', 'ignore'), 's' elif fmt[-1] == 'U': # unicode normalized value, fmt = unicodedata.normalize( @@ -1310,7 +1293,7 @@ class YoutubeDL(object): if date is not None: dateRange = self.params.get('daterange', DateRange()) if date not in dateRange: - return '%s upload date is not in range %s' % (date_from_str(date).isoformat(), dateRange) + return f'{date_from_str(date).isoformat()} upload date is not in range {dateRange}' view_count = info_dict.get('view_count') if view_count is not None: min_views = self.params.get('min_views') @@ -1329,7 +1312,16 @@ class YoutubeDL(object): except TypeError: # For backward compatibility ret = None if incomplete else match_filter(info_dict) - if ret is not None: + if ret is NO_DEFAULT: + while True: + filename = self._format_screen(self.prepare_filename(info_dict), self.Styles.FILENAME) + reply = input(self._format_screen( + f'Download "{filename}"? (Y/n): ', self.Styles.EMPHASIS)).lower().strip() + if reply in {'y', ''}: + return None + elif reply == 'n': + return f'Skipping {video_title}' + elif ret is not None: return ret return None @@ -1731,6 +1723,7 @@ class YoutubeDL(object): entries.append(entry) try: if entry is not None: + # TODO: Add auto-generated fields self._match_entry(entry, incomplete=True, silent=True) except (ExistingVideoReached, RejectedVideoReached): broken = True @@ -1774,21 +1767,22 @@ class YoutubeDL(object): x_forwarded_for = ie_result.get('__x_forwarded_for_ip') - self.to_screen('[%s] playlist %s: %s' % (ie_result['extractor'], playlist, msg % n_entries)) + self.to_screen(f'[{ie_result["extractor"]}] playlist {playlist}: {msg % n_entries}') failures = 0 max_failures = self.params.get('skip_playlist_after_errors') or float('inf') for i, entry_tuple in enumerate(entries, 1): playlist_index, entry = entry_tuple if 'playlist-index' in self.params.get('compat_opts', []): playlist_index = playlistitems[i - 1] if playlistitems else i + playliststart - 1 - self.to_screen('[download] Downloading video %s of %s' % (i, n_entries)) + self.to_screen('[download] Downloading video %s of %s' % ( + self._format_screen(i, self.Styles.ID), self._format_screen(n_entries, self.Styles.EMPHASIS))) # This __x_forwarded_for_ip thing is a bit ugly but requires # minimal changes if x_forwarded_for: entry['__x_forwarded_for_ip'] = x_forwarded_for extra = { 'n_entries': n_entries, - '_last_playlist_index': max(playlistitems) if playlistitems else (playlistend or n_entries), + '__last_playlist_index': max(playlistitems) if playlistitems else (playlistend or n_entries), 'playlist_count': ie_result.get('playlist_count'), 'playlist_index': playlist_index, 'playlist_autonumber': i, @@ -1949,7 +1943,7 @@ class YoutubeDL(object): def syntax_error(note, start): message = ( 'Invalid format specification: ' - '{0}\n\t{1}\n\t{2}^'.format(note, format_spec, ' ' * start[1])) + '{}\n\t{}\n\t{}^'.format(note, format_spec, ' ' * start[1])) return SyntaxError(message) PICKFIRST = 'PICKFIRST' @@ -2053,7 +2047,7 @@ class YoutubeDL(object): raise syntax_error('Expected a selector', start) current_selector = FormatSelector(MERGE, (selector_1, selector_2), []) else: - raise syntax_error('Operator not recognized: "{0}"'.format(string), start) + raise syntax_error(f'Operator not recognized: "{string}"', start) elif type == tokenize.ENDMARKER: break if current_selector: @@ -2247,13 +2241,13 @@ class YoutubeDL(object): return selector_function(ctx_copy) return final_selector - stream = io.BytesIO(format_spec.encode('utf-8')) + stream = io.BytesIO(format_spec.encode()) try: - tokens = list(_remove_unused_ops(compat_tokenize_tokenize(stream.readline))) + tokens = list(_remove_unused_ops(tokenize.tokenize(stream.readline))) except tokenize.TokenError: raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec))) - class TokenIterator(object): + class TokenIterator: def __init__(self, tokens): self.tokens = tokens self.counter = 0 @@ -2339,13 +2333,17 @@ class YoutubeDL(object): # TODO: move sanitization here if is_video: # playlists are allowed to lack "title" - info_dict['fulltitle'] = info_dict.get('title') - if 'title' not in info_dict: + title = info_dict.get('title', NO_DEFAULT) + if title is NO_DEFAULT: raise ExtractorError('Missing "title" field in extractor result', video_id=info_dict['id'], ie=info_dict['extractor']) - elif not info_dict.get('title'): - self.report_warning('Extractor failed to obtain "title". Creating a generic title instead') - info_dict['title'] = f'{info_dict["extractor"]} video #{info_dict["id"]}' + info_dict['fulltitle'] = title + if not title: + if title == '': + self.write_debug('Extractor gave empty title. Creating a generic title') + else: + self.report_warning('Extractor failed to obtain "title". Creating a generic title instead') + info_dict['title'] = f'{info_dict["extractor"].replace(":", "-")} video #{info_dict["id"]}' if info_dict.get('duration') is not None: info_dict['duration_string'] = formatSeconds(info_dict['duration']) @@ -2358,11 +2356,9 @@ class YoutubeDL(object): if info_dict.get(date_key) is None and info_dict.get(ts_key) is not None: # Working around out-of-range timestamp values (e.g. negative ones on Windows, # see http://bugs.python.org/issue1646728) - try: + with contextlib.suppress(ValueError, OverflowError, OSError): upload_date = datetime.datetime.utcfromtimestamp(info_dict[ts_key]) info_dict[date_key] = upload_date.strftime('%Y%m%d') - except (ValueError, OverflowError, OSError): - pass live_keys = ('is_live', 'was_live') live_status = info_dict.get('live_status') @@ -2411,7 +2407,7 @@ class YoutubeDL(object): def sanitize_numeric_fields(info): for numeric_field in self._NUMERIC_FIELDS: field = info.get(numeric_field) - if field is None or isinstance(field, compat_numeric_types): + if field is None or isinstance(field, (int, float)): continue report_force_conversion(numeric_field, 'numeric', 'int') info[numeric_field] = int_or_none(field) @@ -2462,10 +2458,11 @@ class YoutubeDL(object): else: formats = info_dict['formats'] - info_dict['__has_drm'] = any(f.get('has_drm') for f in formats) + # or None ensures --clean-infojson removes it + info_dict['_has_drm'] = any(f.get('has_drm') for f in formats) or None if not self.params.get('allow_unplayable_formats'): formats = [f for f in formats if not f.get('has_drm')] - if info_dict['__has_drm'] and all( + if info_dict['_has_drm'] and all( f.get('acodec') == f.get('vcodec') == 'none' for f in formats): self.report_warning( 'This video is DRM protected and only images are available for download. ' @@ -2653,7 +2650,7 @@ class YoutubeDL(object): if max_downloads_reached: break - write_archive = set(f.get('__write_download_archive', False) for f in formats_to_download) + write_archive = {f.get('__write_download_archive', False) for f in formats_to_download} assert write_archive.issubset({True, False, 'ignore'}) if True in write_archive and False not in write_archive: self.record_download_archive(info_dict) @@ -2721,7 +2718,7 @@ class YoutubeDL(object): for lang in requested_langs: formats = available_subs.get(lang) if formats is None: - self.report_warning('%s subtitles not available for %s' % (lang, video_id)) + self.report_warning(f'{lang} subtitles not available for {video_id}') continue for ext in formats_preference: if ext == 'best': @@ -2764,7 +2761,7 @@ class YoutubeDL(object): tmpl = format_tmpl(tmpl) self.to_screen(f'[info] Writing {tmpl!r} to: {filename}') if self._ensure_dir_exists(filename): - with io.open(filename, 'a', encoding='utf-8') as f: + with open(filename, 'a', encoding='utf-8') as f: f.write(self.evaluate_outtmpl(tmpl, info_copy) + '\n') def __forced_printings(self, info_dict, filename, incomplete): @@ -2834,7 +2831,7 @@ class YoutubeDL(object): urls = '", "'.join( (f['url'].split(',')[0] + ',<data>' if f['url'].startswith('data:') else f['url']) for f in info.get('requested_formats', []) or [info]) - self.write_debug('Invoking downloader on "%s"' % urls) + self.write_debug(f'Invoking {fd.FD_NAME} downloader on "{urls}"') # Note: Ideally info should be a deep-copied so that hooks cannot modify it. # But it may contain objects that are not deep-copyable @@ -2929,11 +2926,11 @@ class YoutubeDL(object): else: try: self.to_screen('[info] Writing video annotations to: ' + annofn) - with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile: + with open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile: annofile.write(info_dict['annotations']) except (KeyError, TypeError): self.report_warning('There are no annotations to write.') - except (OSError, IOError): + except OSError: self.report_error('Cannot write annotations file: ' + annofn) return @@ -2952,13 +2949,13 @@ class YoutubeDL(object): return True try: self.to_screen(f'[info] Writing internet shortcut (.{link_type}) to: {linkfn}') - with io.open(encodeFilename(to_high_limit_path(linkfn)), 'w', encoding='utf-8', - newline='\r\n' if link_type == 'url' else '\n') as linkfile: + with open(encodeFilename(to_high_limit_path(linkfn)), 'w', encoding='utf-8', + newline='\r\n' if link_type == 'url' else '\n') as linkfile: template_vars = {'url': url} if link_type == 'desktop': template_vars['filename'] = linkfn[:-(len(link_type) + 1)] linkfile.write(LINK_TEMPLATES[link_type] % template_vars) - except (OSError, IOError): + except OSError: self.report_error(f'Cannot write internet shortcut {linkfn}') return False return True @@ -3023,10 +3020,10 @@ class YoutubeDL(object): return False # Check extension - exts = set(format.get('ext') for format in formats) + exts = {format.get('ext') for format in formats} COMPATIBLE_EXTS = ( - set(('mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma')), - set(('webm',)), + {'mp3', 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'ismv', 'isma'}, + {'webm'}, ) for ext_sets in COMPATIBLE_EXTS: if ext_sets.issuperset(exts): @@ -3059,7 +3056,7 @@ class YoutubeDL(object): os.path.splitext(filename)[0] if filename_real_ext in (old_ext, new_ext) else filename) - return '%s.%s' % (filename_wo_ext, ext) + return f'{filename_wo_ext}.{ext}' # Ensure filename always has a correct extension for successful merge full_filename = correct_ext(full_filename) @@ -3144,10 +3141,10 @@ class YoutubeDL(object): except network_exceptions as err: self.report_error('unable to download video data: %s' % error_to_compat_str(err)) return - except (OSError, IOError) as err: + except OSError as err: raise UnavailableVideoError(err) except (ContentTooShortError, ) as err: - self.report_error('content too short (expected %s bytes and served %s)' % (err.expected, err.downloaded)) + self.report_error(f'content too short (expected {err.expected} bytes and served {err.downloaded})') return if success and full_filename != '-': @@ -3160,16 +3157,16 @@ class YoutubeDL(object): if fixup_policy in ('ignore', 'never'): return elif fixup_policy == 'warn': - do_fixup = False + do_fixup = 'warn' elif fixup_policy != 'force': assert fixup_policy in ('detect_or_warn', None) if not info_dict.get('__real_download'): do_fixup = False def ffmpeg_fixup(cndn, msg, cls): - if not cndn: + if not (do_fixup and cndn): return - if not do_fixup: + elif do_fixup == 'warn': self.report_warning(f'{vid}: {msg}') return pp = cls(self) @@ -3192,17 +3189,18 @@ class YoutubeDL(object): FFmpegFixupM4aPP) downloader = get_suitable_downloader(info_dict, self.params) if 'protocol' in info_dict else None - downloader = downloader.__name__ if downloader else None + downloader = downloader.FD_NAME if downloader else None if info_dict.get('requested_formats') is None: # Not necessary if doing merger - ffmpeg_fixup(downloader == 'HlsFD', + ffmpeg_fixup(downloader == 'hlsnative' and not self.params.get('hls_use_mpegts') + or info_dict.get('is_live') and self.params.get('hls_use_mpegts') is None, 'Possible MPEG-TS in MP4 container or malformed AAC timestamps', FFmpegFixupM3u8PP) ffmpeg_fixup(info_dict.get('is_live') and downloader == 'DashSegmentsFD', 'Possible duplicate MOOV atoms', FFmpegFixupDuplicateMoovPP) - ffmpeg_fixup(downloader == 'WebSocketFragmentFD', 'Malformed timestamps detected', FFmpegFixupTimestampPP) - ffmpeg_fixup(downloader == 'WebSocketFragmentFD', 'Malformed duration detected', FFmpegFixupDurationPP) + ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed timestamps detected', FFmpegFixupTimestampPP) + ffmpeg_fixup(downloader == 'web_socket_fragment', 'Malformed duration detected', FFmpegFixupDurationPP) fixup() try: @@ -3235,9 +3233,6 @@ class YoutubeDL(object): res = func(*args, **kwargs) except UnavailableVideoError as e: self.report_error(e) - except MaxDownloadsReached as e: - self.to_screen(f'[info] {e}') - raise except DownloadCancelled as e: self.to_screen(f'[info] {e}') if not self.params.get('break_per_url'): @@ -3292,9 +3287,9 @@ class YoutubeDL(object): info_dict.setdefault('_type', 'video') if remove_private_keys: - reject = lambda k, v: v is None or (k.startswith('_') and k != '_type') or k in { + reject = lambda k, v: v is None or k.startswith('__') or k in { 'requested_downloads', 'requested_formats', 'requested_subtitles', 'requested_entries', - 'entries', 'filepath', 'infojson_filename', 'original_url', 'playlist_autonumber', + 'entries', 'filepath', '_filename', 'infojson_filename', 'original_url', 'playlist_autonumber', } else: reject = lambda k, v: False @@ -3316,6 +3311,17 @@ class YoutubeDL(object): ''' Alias of sanitize_info for backward compatibility ''' return YoutubeDL.sanitize_info(info_dict, actually_filter) + def _delete_downloaded_files(self, *files_to_delete, info={}, msg=None): + for filename in set(filter(None, files_to_delete)): + if msg: + self.to_screen(msg % filename) + try: + os.remove(filename) + except OSError: + self.report_warning(f'Unable to delete file {filename}') + if filename in info.get('__files_to_move', []): # NB: Delete even if None + del info['__files_to_move'][filename] + @staticmethod def post_extract(info_dict): def actual_post_extract(info_dict): @@ -3348,14 +3354,8 @@ class YoutubeDL(object): for f in files_to_delete: infodict['__files_to_move'].setdefault(f, '') else: - for old_filename in set(files_to_delete): - self.to_screen('Deleting original file %s (pass -k to keep)' % old_filename) - try: - os.remove(encodeFilename(old_filename)) - except (IOError, OSError): - self.report_warning('Unable to remove downloaded original file') - if old_filename in infodict['__files_to_move']: - del infodict['__files_to_move'][old_filename] + self._delete_downloaded_files( + *files_to_delete, info=infodict, msg='Deleting original file %s (pass -k to keep)') return infodict def run_all_pps(self, key, info, *, additional_pps=None): @@ -3397,7 +3397,7 @@ class YoutubeDL(object): break else: return - return '%s %s' % (extractor.lower(), video_id) + return f'{extractor.lower()} {video_id}' def in_download_archive(self, info_dict): fn = self.params.get('download_archive') @@ -3594,7 +3594,7 @@ class YoutubeDL(object): def urlopen(self, req): """ Start an HTTP download """ - if isinstance(req, compat_basestring): + if isinstance(req, str): req = sanitized_Request(req) return self._opener.open(req, timeout=self._socket_timeout) @@ -3605,7 +3605,7 @@ class YoutubeDL(object): def get_encoding(stream): ret = str(getattr(stream, 'encoding', 'missing (%s)' % type(stream).__name__)) if not supports_terminal_sequences(stream): - from .compat import WINDOWS_VT_MODE + from .compat import WINDOWS_VT_MODE # Must be imported locally ret += ' (No VT)' if WINDOWS_VT_MODE is False else ' (No ANSI)' return ret @@ -3652,10 +3652,8 @@ class YoutubeDL(object): if re.match('[0-9a-f]+', out): write_debug('Git HEAD: %s' % out) except Exception: - try: + with contextlib.suppress(Exception): sys.exc_clear() - except Exception: - pass def python_implementation(): impl_name = platform.python_implementation() @@ -3672,7 +3670,7 @@ class YoutubeDL(object): exe_versions, ffmpeg_features = FFmpegPostProcessor.get_versions_and_features(self) ffmpeg_features = {key for key, val in ffmpeg_features.items() if val} if ffmpeg_features: - exe_versions['ffmpeg'] += ' (%s)' % ','.join(ffmpeg_features) + exe_versions['ffmpeg'] += ' (%s)' % ','.join(sorted(ffmpeg_features)) exe_versions['rtmpdump'] = rtmpdump_version() exe_versions['phantomjs'] = PhantomJSwrapper._version() @@ -3681,20 +3679,12 @@ class YoutubeDL(object): ) or 'none' write_debug('exe versions: %s' % exe_str) - from .downloader.websocket import has_websockets - from .postprocessor.embedthumbnail import has_mutagen - from .cookies import SQLITE_AVAILABLE, SECRETSTORAGE_AVAILABLE - - lib_str = join_nonempty( - compat_brotli and compat_brotli.__name__, - has_certifi and 'certifi', - compat_pycrypto_AES and compat_pycrypto_AES.__name__.split('.')[0], - SECRETSTORAGE_AVAILABLE and 'secretstorage', - has_mutagen and 'mutagen', - SQLITE_AVAILABLE and 'sqlite', - has_websockets and 'websockets', - delim=', ') or 'none' - write_debug('Optional libraries: %s' % lib_str) + from .compat.compat_utils import get_package_info + from .dependencies import available_dependencies + + write_debug('Optional libraries: %s' % (', '.join(sorted({ + join_nonempty(*get_package_info(m)) for m in available_dependencies.values() + })) or 'none')) self._setup_opener() proxy_map = {} @@ -3705,10 +3695,10 @@ class YoutubeDL(object): # Not implemented if False and self.params.get('call_home'): - ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode('utf-8') + ipaddr = self.urlopen('https://yt-dl.org/ip').read().decode() write_debug('Public IP address: %s' % ipaddr) latest_version = self.urlopen( - 'https://yt-dl.org/latest/version').read().decode('utf-8') + 'https://yt-dl.org/latest/version').read().decode() if version_tuple(latest_version) > version_tuple(__version__): self.report_warning( 'You are using an outdated version (newest version: %s)! ' @@ -3744,7 +3734,7 @@ class YoutubeDL(object): https_handler = make_HTTPS_handler(self.params, debuglevel=debuglevel) ydlh = YoutubeDLHandler(self.params, debuglevel=debuglevel) redirect_handler = YoutubeDLRedirectHandler() - data_handler = compat_urllib_request_DataHandler() + data_handler = urllib.request.DataHandler() # When passing our own FileHandler instance, build_opener won't add the # default FileHandler and allows us to disable the file protocol, which @@ -3800,7 +3790,7 @@ class YoutubeDL(object): try: write_json_file(self.sanitize_info(ie_result, self.params.get('clean_infojson', True)), infofn) return True - except (OSError, IOError): + except OSError: self.report_error(f'Cannot write {label} metadata to JSON file {infofn}') return None @@ -3821,9 +3811,9 @@ class YoutubeDL(object): else: try: self.to_screen(f'[info] Writing {label} description to: {descfn}') - with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile: + with open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile: descfile.write(ie_result['description']) - except (OSError, IOError): + except OSError: self.report_error(f'Cannot write {label} description file {descfn}') return None return True @@ -3857,12 +3847,12 @@ class YoutubeDL(object): try: # Use newline='' to prevent conversion of newline characters # See https://github.com/ytdl-org/youtube-dl/issues/10268 - with io.open(sub_filename, 'w', encoding='utf-8', newline='') as subfile: + with open(sub_filename, 'w', encoding='utf-8', newline='') as subfile: subfile.write(sub_info['data']) sub_info['filepath'] = sub_filename ret.append((sub_filename, sub_filename_final)) continue - except (OSError, IOError): + except OSError: self.report_error(f'Cannot write video subtitles file {sub_filename}') return None |